| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
High Performance Steels
|
| Presentation Title |
Physics-Coupled Data-Driven Design of Advanced Alumina-Forming Austenitic Stainless Steel |
| Author(s) |
Dongwon Shin, Sun Yong Kwon, Peng Jian, Yukinori Yamamoto, Michael Brady, James Allen Haynes |
| On-Site Speaker (Planned) |
Dongwon Shin |
| Abstract Scope |
We present a materials design loop that uses physics-guided machine learning (ML) models to discover new alloy chemistries with improved properties by identifying a high-temperature alumina-forming austenitic stainless steel with enhanced creep resistance. Our ML models were trained using a well-curated experimental dataset supplemented with synthetic microstructural information from computational thermodynamics. We have explored a large number of hypothetical alloys in a high-dimensional composition space and predict their creep properties. We used uncertainties from the ML training as thresholds to truncate predicted results and identify alloys with improved or deteriorated creep. We utilize probability density distribution analysis to determine elemental compositions for further virtual and experimental validations. This research was supported by the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for the U.S. Department of Energy and Vehicle Technologies Office. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
ICME, Iron and Steel, |